BACKGROUND OF THE INVENTION Blood coagulation is a process consisting of a complex interaction of various blood components, or factors, which eventually gives rise to a fibrin clot. Generally, the blood com- ponents which participate in what has been referred to as the coagulation"cascade"are pro- enzymes or zymogens, enzymatically inactive proteins which are converted to proteolytic en- zymes by the action of an activator, itself an activated clotting factor. Coagulation factors that have undergone such a conversion and generally referred to as"active factors, "and are des- ignated by the addition of a lower case"a"suffix (e. g., FactorVlla).

Activated Factor X ("Xa") is required to convert prothrombin to thrombin, which then converts fibrinogen to fibrin as a final stage in forming a fibrin clot. There are two systems, or pathways, that promote the activation of Factor X. The"intrinsic pathway"refers to those reactions that lead to thrombin formation through utilization of factors present only in plasma.

A series of protease-mediated activations ultimately generates Factor IXa which, in conjunction with Factor VI Ila, cleaves Factor X into Xa. An identical proteolysis is effected by FVlla and its co-factor, TF, in the"extrinsic pathway"of blood coagulation. TF is a membrane bound protein and does not normally circulate in an active form in plasma. Upon vessel disruption, however, TF can complex with FVlla to catalyze Factor X activation or Factor lX activation in the presence of Ca2+ and phospholipid. While the relative importance of the two coagulation pathways in hemostasis is unclear, Factor VII and TF have been found to play a pivotal role in the initiation of blood coagulation.

It is often necessary to selectively block the coagulation cascade in a patient. Anticoagulants such as heparin, coumarin, derivatives of coumarin, indandione derivatives, or other agents

may be used, for example, during kidney dialysis, or to treat deep vein thrombosis, disseminated intravascular coagulation (DIC), and a host of other medical disorders. For example, heparin treatment or extracorporeal treatment with citrate ion may be used in dialysis to prevent coagulation during the course of treatment. Heparin is also used in preventing deep vein thrombosis in patients undergoing surgery.

Treatment with heparin and other anticoagulants may, however, have undesirable side effects. Available anticoagulants generally act throughout the body, rather than acting specifically at site of injury. Heparin, for example, may cause heavy bleeding. Furthermore, with a half-life of approximately 80 minutes, heparin is rapidly cleared from the blood, necessitating frequent administration. Because heparin acts as a cofactor for antithrombin III (ATIII), and ATIII is rapidly depleted in DIC treatment, it is often difficult to maintain the proper heparin dosage, necessitating continuous monitoring of ATIII and heparin levels.

Heparin is also ineffective if ATIII depletion is extreme. Further, prolonged use of heparin may also increase platelet aggregation and reduce platelet count, and has been implicated in the development of heparin-induced thrombocytopenia. Indandione derivatives may also have toxic side effects. In addition to the anticoagulants briefly described above, several naturally occurring proteins have been found to have anticoagulant activity. Also, ATIII has been proposed as a therapeutic anticoagulant.

International Application No. WO 92/15686 relates to inactivated Factor Vlla for inhibiting blood coagulation.

Antibodies are specific immunoglobulin (Ig) polypeptides produced by the vertebrate immune system in response to challenges by foreign proteins, glycoproteins, cells, or other antigenic foreign substances. The sequence of events which permits the organism to overcome invasion by foreign cells or to rid the system of foreign substances is at least partially understood. An important part of this process is the manufacture of antibodies which bind specifically to a particular foreign substance. The binding specificity of such polypeptides to a particular antigen is highly refined, and the multitude of specificities capable of being generated by the individual vertebrate is remarkable in its complexity and variability. Millions of antigens are capable of eliciting antibody responses, each antibody almost exclusively directed to the particular antigen which elicited it.

Two major sources of vertebrate antibodies are presently utilized, generation in situ by the mammalian B lymphocytes, and generation in cell culture by B-cell hybrids. Antibodies are generated in situ as a result of the differentiation of immature B lymphocytes into plasma cells, which occurs in response to stimulation by specific antigens. In the undifferentiated B cells, the portions of DNA coding for the various regions on the immunoglobulin chains are separated in the genomic DNA. The sequences are assembled sequentially prior to

expression. The resulting rearranged gene is capable of expression in the mature B lymphocyte to produce the desired antibody. However, even when a particular mammal is exposed to only a single antigen a uniform population of antibodies does not result. The in situ immune response to any particular antigen is defined by the mosaic of responses to the various determinants which are present on the antigen. Each subset of homologous antibodies is contributed by a single population of B cells, hence in situ generation of antibodies is"polyclonal".

This limited but inherent heterogeneity has been overcome in numerous particular cases by use of hybridoma technology to create"monoclonal"antibodies in cell cultures by B cell hybridomas.

In this process, the relatively short-lived, or mortal, splenocytes or lymphocytes from a mammal which has been injected with antigen are fused with an immortal tumor cell line, thus producing hybrid cells or"hybridomas"which are both immortal and capable of producing the genetically coded antibody of the B cell. The hybrids thus formed are segregated into single genetic strains by selection, dilution, and regrowth, and each strain thus represents a single genetic line. They therefore, produce antibodies which are assured to be homogeneous against a desired antigen. These antibodies, referencing their pure genetic parentage, are called"monoclonal".

Monoclonal antibodies with mono-specificity have greatly influenced immunology, and their usefulness has already been demonstrated in such sciences as biology, pharmacology, biochemistry and others. Such monoclonal antibodies have found widespread use not only as diagnostics reagents, but also therapeutically (see, for example, Ritz and Schlossman, Blood, 59: 1-11, (1982)).

Monoclonal antibodies produced by hybridomas, while theoretically effective as discussed above and clearly preferable to polyclonal antibodies because of their specificity, suffer from an important disadvantage. In many applications, the use of monoclonal antibodies produced in non-human animals is severely restricted where the monoclonal antibodies are to be used in humans. Repeated injections of a"foreign"antibody in humans, such as a mouse antibody, may lead to harmful hypersensitivity reactions. Such a non-human derived monoclonal antibody, when injected into humans, causes an anti-nonhuman antibody response.

Therapeutic use of mouse Mabs against TF is known from U. S. patent no. 6,001, 978 and 5,223, 427.

International Application No. WO 99/51743 relates to human/mouse chimera monoclonal antibodies directed against human TF.

European patent application No. 833911 relates to CDR-grafted antibodies against human

TF.

Presta L. et al., Thrombosis and Haemostasis, Vol. 85 (3) pp. 379-389 (2001) relates to humanized antibody against TF inhibiting plasma clotting and TF/FVlla-mediated activation of factors IX and X. There is still a need in the art for improved compositions having anticoagulant activity which can be administered at relatively low doses and do not produce the undesirable side effects associated with traditional anticoagulant compositions. The present invention fulfils this need by providing anticoagulants that do not have the side effects associated with the traditional antibodies with non-human sequences, they act specifically at sites of injury, and further provides other related advantages. Furthermore the present invention provides compounds, which acts to inhibit the cellular functions of TF, which is implicated in conditions like sepsis, inflammation, atherosclerosis, restenosis, or cancer.

These and other advantages of the invention, as well as additional inventive features, will be apparent from the description of the invention provided herein.

SUMMARY OF THE INVENTION In one aspect, the invention provides non-immunogenic high affinity humanized antibodies against human TF, which antibodies inhibit the binding of coagulation factor VIINlla. In another aspect, the invention provides novel and useful methods for selection of therapeutically effective humanized antibodies against human TF. In a more particular aspect, the present invention relates to an isolated humanized antibody, that immunoreacts with a first epitope present on human TF and inhibits the binding of human coagulation factor Vlla to human TF. In other aspects, the invention relates to methods of using these and other TF antibodies, methods of producing such antibodies, cells for producing such antibodies, and related compositions.

Detailed DESCRIPTION OF THE INVENTION The terms"humanized antibodies"and"humanized antibody", as used herein, refer to recombinant antibodies in which complementarity determining region (CDR) sequences or variants thereof derived from the germline of another mammalian species, such as a mouse, have been grafted into human framework sequences, e. g. the so-called humanized antibodies. In one aspect of the invention the isolated humanized monoclonal antibody comprises CDR amino acid sequences derived from a monoclonal antibody. In a further aspect of the invention the isolated humanized monoclonal antibody has CDR amino acid sequences derived from a mouse monoclonal antibody.

The term"monoclonal antibody"as used herein, refers to a homogeneous population of immunoglobulins, i. e. the individual molecules of the antibody population are identical except for naturally occurring mutations. Antibodies are normally synthesized by lymphoid cells

derived from B lymphocytes of bone marrow. Lymphocytes derived from the same clone produce immunoglobulin of a single amino acid sequence. Lymphocytes can not be directly cultured over long periods of time to produce substantial amounts of their specific antibody.

However, Kohler et al., 1975, Nature, 256: 495, demonstrated that a process of somatic cell fusion, specifically between a lymphocyte and a myeloma cell, could yield hybridoma cells which grow in culture and produce a specific antibody called a"monoclonal antibody".

Myeloma cells are lymphocyte tumor cells which, depending upon the cell strain, frequently produce an antibody themselves, although"non-producing"strains are known.

The term"recombinant antibody", as used herein, is intended to include all antibodies that are prepared, expressed, created or isolated by recombinant means, such as antibodies expressed using a recombinant expression vector transfected into a host cell (described further in Section II, below), antibodies isolated from a recombinant combinatorial antibody library (described further in Section III, below), antibodies isolated from an animal (e. g. , a mouse) that is transgenic for human immunoglobulin genes (see e. g., Taylor, L. D. , et al.

(1992) Nucl. Acids Res. 20: 6287-6295) and/or antibodies prepared, expressed, created or isolated by any other means that involves splicing of human immunoglobulin gene sequences to other DNA sequences. Such recombinant human antibodies have variable and constant regions derived from human germline immunoglobulin sequences. In certain aspects of the invention, however, such recombinant human antibodies are subjected to in vitro mutagenesis (or, when an animal transgenic for human Ig sequences is used, in vivo somatic mutagenesis) and thus the amino acid sequences of the VH and VL regions of the recombinant antibodies may be sequences that, while derived from and related to human germline VH and VL sequences, may not naturally exist within the human antibody germline repertoire in vivo.

The terms"human tissue factor"or"human TF"as used herein, refers to the full length polypeptide receptor comprising the amino acid sequence 1-263 of native human tissue factor.

The term"antibody", as used herein, refers to immunoglobulin molecules and fragments thereof, that have the ability to specifically bind to an antigen (e. g. , human TF). Full-length antibodies comprises four polypeptide chains, two heavy (H) chains and two light (L) chains inter-connected by disulfide bonds. Each heavy chain is comprised of a heavy chain variable region (abbreviated herein as HCVR or VH) and a heavy chain constant region. The heavy chain constant region is comprised of three domains, CH1, CH2 and CH3. Each light chain is comprised of a light chain variable region (abbreviated herein as LCVR or VL) and a light chain constant region. The light chain constant region is comprised of one domain, CL. The VH and VL regions can be further subdivided into regions of hypervariability, termed

complementarity determining regions (CDR), interspersed with regions that are more conserved, termed framework regions (FR). Each VH and VL is composed of three CDRs and four FRs, arranged from amino-terminus to carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4. Thus, within the definition of an antibody is also one or more fragments of an antibody that retain the ability to specifically bind to an antigen (e. g. , human TF). It has been shown that the antigen-binding function of an antibody can be performed by fragments of a full-length antibody. Examples of binding fragments encompassed within the term"antibody"include (i) a Fab fragment, a monovalent fragment consisting of the VL, VH, CL and CH I domains; (ii) F (ab) 2 and F (ab') 2 fragments, a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting of the VH and CH1 domains; (iv) a Fv fragment consisting of the VL and VH domains of a single arm of an antibody, (v) a dAb fragment (see e. g. Ward et al., (1989) Nature 341: 544-546), which consists of a VH domain; and (vi) an isolated complementarity determining region (CDR). Furthermore, although the two domains of the Fv fragment, VL and VH, are coded for by separate genes, they can be joined, using recombinant methods, by a synthetic linker that enables them to be made as a single protein chain in which the VL and VH regions pair to form monovalent molecules (known as single chain Fv (scFv); see e. g. , Bird et al. (1988) Science 242: 423-426: and Huston et al. (1988) Proc. Natl. Acad. Sci. USA 85: 5879-5883). Such single chain antibodies are also intended to be encompassed within the term"antibody". Other forms of single chain antibodies, such as diabodies are also encompassed by the term"antibody.". Diabodies are bivalent and bispecific antibodies in which VH and VL domains are expressed on a single polypeptide chain, but are separated by a linker that is too short to allow for pairing between the two domains on the same chain, thereby forcing the domains to pair with complementary domains of another chain so as to create two antigen binding sites (see e. g., Holliger, P. , et al. (1993) Proc. Natl. Acad. Sci. USA 90: 6444-6448; Poijak, R. J. , et al. (1994) Structure 2: 1121-1123). It is understood that human TF may have one or more antigenic determinants comprising (1) peptide antigenic determinants which consist of single peptide chains within human TF, (2) conformational antigenic determinants which consist of more than one spatially contiguous peptide chains whose respective amino acid sequences are located disjointedly along the human TF polypeptide sequence; and (3) post-translational antigenic determinants which consist, either in whole or part, of molecular structures covalently attached to human TF after translation, such as carbohydrate groups.

An"isolated humanized antibody", as used herein, refers to a humanized antibody that is substantially free of other antibodies having different antigenic specificities (e. g. , an isolated antibody that specifically binds human TF is substantially free of antibodies that specifically

bind antigens other than human TF). An isolated antibody that specifically binds human TF may, however, have cross-reactivity to other antigens, such as TF molecules from other species (discussed in further detail below). Moreover, an isolated antibody may be substantially free of other cellular material and/or chemicals.

The term"epitope"as used herein means any antigenic determinant on an antigen to which the antibody binds. Epitopic determinants usually consist of chemically active surface groupings of molecules such as amino acids or sugar side chains and usually have specific three dimensional structural characteristics, as well as specific charge characteristics.

The terms"immunoreacts"or"immunoreacting", as used herein, means any binding of an antibody to its epitope with a dissociation constant Kd lower than 104 M. The terms "immunoreacts"or"immunoreacting"are used where appropriate interchangeably with the phrase"specifically bind".

The term"inhibits", as used herein, means any reduction compared to a reference. As an example, an antibody, that"inhibits"the binding of human coagulation factor Vlla to human TF means any antibody, that reduces the ability of human coagulation factor Vlla to bind human TF compared to the ability of human coagulation factor Vlla to bind human TF in the absense of the antibody.

The term"affinity", as used herein, means the strength of the binding of an antibody to an epitope. The affinity of an antibody is measured by the dissociation constant Kd, defined as [Ab] x [Ag] / [Ab-Ag] where [Ab-Ag] is the molar concentration of the antibody-antigen complex, [Ab] is the molar concentration of the unbound antibody and [Ag] is the molar concentration of the unbound antigen. The affinity constant Ka is defined by 1/Kd. Preferred methods for determining antibody specificity and affinity by competitive inhibition can be found in Harlow, et al., Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N. Y. , 1988), Colligan et al., eds. , Current Protocols in Immunology, Greene Publishing Assoc. and Wiley Interscience, N. Y. , (1992,1993), and Muller, Meth. Enzymol. 92: 589-601 (1983), which references are entirely incorporated herein by reference.

In another aspect, the invention relates to a pharmaceutical composition comprising a therapeutically effective amount of a humanized antibody, that immunoreacts with a first epitope present on human TF and inhibits the binding of human coagulation factor Vlla to human TF.

The phrase"a therapheutically effective amount"is the effective dose to be determined by a qualified practitioner, who may titrate dosages to achieve the desired response. Factors for consideration of dose will include potency, bioavailability, desired pharmacokinetic/pharmacodynamic profiles, condition of treatment (e. g. trauma,

inflammation, septic chock), patient-related factors (e. g. weight, health, age, etc. ), presence of co-administered medications, time of administration, and/or other factors known to a medical practitioner. The dosage of a humanized antibody against TF administered to a patient will vary with the type and severity of the condition to be treated, but is generally in the range of 0.1-5. 0 mg/kg body weight.

The term"subject"as used herein means any animal, in particular mammals, such as humans, and may, where appropriate, be used interchangeably with the term"patient".

In a another aspect, the invention relates to a composition comprising a humanized antibody, that immunoreacts with a first epitope present on human TF and inhibits the binding of human coagulation factor Vlla to human TF.

In a further aspect, the invention relates to a method for treatment of a FVlla/TF related disorder in a human, which method comprises administering to the human a therapeutical effective amount of a humanized antibody, that immunoreacts with a first epitope present on human TF and inhibits the binding of human coagulation factor Vlla to human TF.

"Treatment"means the administration of an amount of antibodies of the invention and/or a compound of the invention with the effect of reducing the intensity of, delaying the onset of, and/or reducing the risk of, and preferably preventing, the occurrence of one or more physiological symptoms associated with a disease state or reducing and/or eliminating one or more detectable physiological conditions or symptoms associated with a disease state present in a subject (and, preferably, eliminating all conditions associated with the disease state). Humanized TF antibodies and compositions provided by the invention can be used in any and all aspects of treating a subject.

The terms"FVlla/TF related disorder"as used herein means a disease or disorder, where TF and FVlla are involved. Examples of FVlla/TF related disorders include thrombotic and/or coagulopathic related diseases, and disorders including inflammatory response and chronic thromboembolic diseases, and disorders associated with fibrin formation including vascular disorders such as deep venous thrombosis, arterial thrombosis, post surgical thrombosis, coronary artery bypass graft (CABG), percutaneous transdermal coronary an- gioplastry (PTCA), stroke, tumor growth, tumor metastasis, angiogenesis-related disease (s), thrombolysis, atherosclerosis, arteriosclerosis e. g. arteriosclerosis following angioplastry, and restenosis (such as restenosis following angioplastry), acute and chronic indications such as inflammation, septic chock, septicemia, hypotension, adult respiratory distress syndrome (ARDS), disseminated intravascular coagulopathy (DIC), pulmonary embolism, platelet deposition, myocardial infarction, or the prophylactic treatment of mammals with atheroscle- rotic vessels at risk for thrombosis, and other diseases or disorders. The FVlla/TF related

disorder is not limited to in vivo coagulopatic disorders such as those named above but in- cludes ex vivo FVlla/TF related processes such as coagulation that may result from the ex- tracorporeal circulation of blood, including blood removed in-line from a patient in such proc- esses as dialysis procedures, blood filtration, or blood bypass during surgery.

The term"Factor Vlla", or"FVlla"means"two chain"activated coagulation factor VII cleaved by specific cleavage at the Arg152-lle153 peptide bond. FVlla, may be purified from blood or produced by recombinant means. It is evident that the practice of the methods described herein is independent of how the purified factor Vlla is derived and, therefore, the applicable aspects of the invention can include the use of any suitable Factor Vlla preparation. In many aspects, hu- man Factor FVlla is preferred.

The term"FVII"means"single chain"coagulation factor Vil.

In a further aspect, the invention relates to a method for preparation of a humanized anti- body, which method comprises preparation of humanized antibodies against human TF, testing antibodies in a FVlla/TF amidolytic assay and selecting humanized antibody which inhibit TF-induced FVlla amidolytic activity, with an IC5ovalue lower than the iCso value of FFR-rFVlla + 100 nM (using 10 nM FVlla in the assay), such as lower than the IC. 50 value of FFR-rFVlla + 40 nM, preferably lower than the IC50 value of FFR-rFVlla + 20 nM, preferably lower than the Iso value of FFR-rFVlla + 10 nM, more preferably lower than the IC50 value of FFR-rFVlla, or testing antibodies in a FVlla competition assay and selecting humanized antibody which compete with FVlla binding, or testing antibodies in a FVlla signalling assay and selecting humanized antibodies which in- hibits FVlla-induced intracellular signalling.

The term"FVlla signalling assay"as used herein means any assay which measures the FVlla-induced intracellular signalling.

The term"FVlla-induced intracellular signalling"as used herein means the intracellular events mediated when FVlla is added to the cell."FVlla-induced intracellular signalling"may be measured by the MAPK signalling assay or a gene expression analysis assay which meas- ures the up-regulation of genes independently selected from the list comprising Fra-1, Id2, and Cyr61.

Unless otherwise indicated, the specific IC50 values referred to with respect to the TF-induced clot assays described herein are those obtained when using normal human plasma.

The term"TF-induced clot assay"as used herein means any assay where clotting time is measured in sample comprising blood plasma and TF. An example of a TF-induced clot as- say is described in Example 1, assay 7.

The term"FXa generation assay"as used herein means any assay where activation of FX is measured in a sample comprising TF, FVlla, FX, calcium and phospholipids. An example of a FXa generation assay is described in Example 1, assay 5.

The term"FVlla/TF amidolytic assay"as used herein means any assay where the amidolytic activity, i. e. cleavage of a small peptide substrate, of FVlla is measured in the presence of TF. An example of a FVlla/TF amidolytic assay is described in Example 1, assay 4.

The term"TF ELISA assay"as used herein means any ELISA assay comprising TF and anti- bodies against TF. Examples of TF ELISA assays are the direct and indirect TF ELISA as- says described in Example 1, assay 1 and 2.

The term"direct TF ELISA assay"as used herein means any TF ELISA assay comprising immobilized TF. An example of a direct TF ELISA assays is described in Example 1, assay 1.

The term"indirect TF ELISA assay"as used herein means any TF ELISA assay, where TF is in solution. Example of direct TF ELISA assays is described in Example 1, assay 2.

In a further aspect, the invention relates to a humanized antibody that immunoreacts with a first epitope present on human TF and inhibits the binding of human coagulation factor Vlla to human TF and are produced by a method that comprises : preparation of humanized antibodies against human TF, testing antibodies in a FVlla/TF amidolytic assay and selecting humanized antibody which inhibit TF-induced FVlla amidolytic activity, with an ICSO value lower than the IC50 value of FFR-rFVlla + 100 nM (using 10 nM FVlla), such as lower than the IC50value of FFR-rFVlla + 40 nM, preferably lower than the In50 value of FFR-rFVlla + 20 nM, preferably lower than the iCso value of FFR-rFVlla + 10 nM, more preferably lower than the IC50value of FFR-rFVlla, or testing antibodies in a FVlla competition assay and selecting humanized antibody which compete with FVlla binding, or testing antibodies in a FVlla signalling assay and selecting humanized antibodies which in- hibits FVlla-induced intracellular signalling.

In a further aspect, the invention relates to a method for preparation of a humanized anti- body, which method comprises a) preparation of immortal cells to secrete humanized antibodies, b) isolation of culture medium from immortal cells comprising produced humanized an- tibodies, c) testing antibodies in an indirect TF ELISA assay comprising TF in solution and se- lecting humanized antibodies that immunoreacts with human TF in solution,

d) testing antibodies in a FVlla competition assay and selecting humanized antibodies which competes with FVlla binding, e) testing antibodies in a FVlla/TF amidolytic assay and selecting humanized antibody which inhibit TF-induced FVlla amidolytic activity, with an IC50 value lower than the IC50 value of FFR-rFVlla + 100 nM (using 10 nm FVlla in the assay), such as lower than the IC50value of FFR-rFVlla + 40 nM, preferably lower than the IC5Ovalue of FFR-rFVlla + 20 nM, preferably lower than the IC50 value of FFR-rFVlla + 10 nM, more preferably lower than the In50 value of FFR-rFVlla, f) testing antibodies in a FXa generation assay and selecting humanized antibody which inhibit FXa generation with an IC50value lower than the IC5ovalue of FFR- rFVlla + 100 nM (using 0.1 nM FVlla in the assay), such as lower than the IC50 value of FFR-rFVlla + 10 nM preferably lower than the IC50 value of FFR-rFVlla + 5 nM, preferably lower than the IC50 value of FFR-rFVlla + 1 nM, more preferably lower than the IC50value of FFR-rFVlla + 0. 1 nM, more preferably lower than the IC50 value of FFR-rFVlla, g) testing antibodies in a TF-induced clot assay and selecting humanized antibodies which inhibits clot formation in this assay with an ICSO value lower than the IC50 value of FFR-rFVlla + 1 nM, such as lower than the IC50 value of FFR-rFVlla + 500 pM, preferably lower than the IC50 value of FFR-rFVlla + 200 pM, preferably lower than the ICgo value of FFR-rFVlla + 100 pM, such as lower than the ICS value of FFR- rFVlla + 50 pM, preferably lower than the IC50 value of FFR-rFVlla + 10 pM, more preferably lower than the IC50value of FFR-rFVlla + 5 pM, more preferably lower than the IC50value of FFR-rFVlla, h) selection and cultivation in a suitable culture medium of the immortal cell producing the selected antibody following step c-g, i) isolation of selected antibody from culture medium of selected immortal cell.

The term"antibody-producing cell"as used herein means any cell capable of pro- ducing an antibody. Included are hybridomas, transfected cell lines and the relatively short- lived, or mortal, splenocytes or lymphocytes from a mammal which has been injected with an antigen.

In a further aspect, the invention relates to a humanized antibody that immunoreacts with a first epitope present on human TF and inhibits the binding of human coagulation factor Vlla to human TF and are produced by a method that comprises: a) preparation of immortal cells to secrete humanized antibodies, b) isolation of culture medium from immortal cells comprising produced antibodies,

c) testing antibodies in an indirect TF ELISA assay comprising TF in solution and se- lecting humanized antibodies that immunoreacts with human TF in solution, d) testing antibodies in a FVlla competition assay and selecting humanized antibodies which competes with FVlla binding, e) testing antibodies in a FVlla/TF amidolytic assay and selecting humanized antibody which inhibit TF-induced FVlla amidolytic activity, with an ICgo value lower than the IC50value of FFR-rFVlla + 100 nM (using 10 nM FVlla in the assay), such as lower than the IC50 value of FFR-rFVlla + 40 nM, preferably lower than the IC5o value of FFR-rFVlla + 20 nM, preferably lower than the IC50 value of FFR-rFVlla + 10 nM, more preferably lower than the ICgo value of FFR-rFVlla, f) testing antibodies in a FXa generation assay and selecting humanized antibody which inhibit FXa generation with an IC50 value lower than the IC50 value of FFR- rFVlla + 100 nM (using 0.1 nM FVlla in the assay), such as lower than the IC50 value of FFR-rFVlla + 10 nM, preferably lower than the IC50 value of FFR-rFVlla + 5 nM, preferably lower than the IC50 value of FFR-rFVlla + 1 nM, more preferably lower than the IC5ovalue of FFR-rFVlla + 0. 1 nM, more preferably lower than the IC50 value of FFR-rFVlla, g) testing antibodies in a TF-induced clot assay and selecting humanized antibodies which inhibits clot formation in this assay with an IC50 value lower than the IC50 value of FFR-rFVlla + 1 nM, such as lower than the IC5o value of FFR-rFVlla + 500 pM, preferably lower than the IC50 value of FFR-rFVlla + 200 pM, preferably lower than the IC50 value of FFR-rFVlla + 100 pM, such as lower than the IC50 value of FFR- rFVlla + 50 pM, preferably lower than the IC50value of FFR-rFVlla + 10 pM, more preferably lower than the In50 value of FFR-rFVlla + 5 pM, more preferably lower than the iCso vaiue of FFR-rFVlla, h) selection and cultivation in a suitable culture medium of the immortal cell producing the selected antibody following step c-g, i) isolation of selected antibody from culture medium of selected immortal cell.

It is to be understood, that the specific ICgo values referred to in the TF-induced clot assay is when using normal human plasma.

In a further aspect, the invention relates to a cell producing humanized antibodies that immunoreacts with a first epitope present on human TF and inhibits the binding of hu- man coagulation factor Vlla to human TF. In one embodiment the cell is a mammal cell. In a further embodiment the cell is isolated from a mouse. In a further embodiment the cell is a hybridoma cell. In a further embodiment the hybridoma cell is obtained by fusion of an anti-

body-producing lymphoid cell with an immortal cell to provide an antibody-producing hybri- doma cell. In a further embodiment the cell is a myeloma cell.

Many myeloma cell lines may be used for the production of humanized antibodies in accordance with the techniques described herein, including, e. g., Mouse myeloma NSO (ECACC 85110503), rat myeloma YB2/0 (ATCC CRL 1662), P3X63-Ag8, P3X63-AG8.653, P3/NS1-Ag4-1 (NS-1), Sp2/0-Agl4 and S194/5. XXO. Bu. 1. The P3X63-Ag8 and NS-1 cell lines have been described by Kohler and Milstein (Eur. J. Immunol. 6: 511 (1976)). Shulman et al. (Nature 276: 269 (1978) ) developed the Sp2/0-Agl4 myeloma line. The S194/5. XXO. Bu. 1 line was reported by Trowbridge (J. Exp. Med. 148: 313 (1979)).

Traditionally humanized or human antibodies are developed by phage display, ribo- some display, RNA display, surface display on a"living organism"or by using transgenic animals or in immune incompetent animals where human immune competent cells are intro- duced. Common for the mentioned display techniques are that the humanized antibodies produced by these techniques have affinities that are below what is normally needed for therapeutic antibodies. One has to affinity mature these humanized antibodies further. One reason is that CDR regions from large libraries are combined with a"naive"human frame- work that has never seen the epitope of interest before, which makes the conformation of this human framework sub optimal. Using transgenic animals where the human Ig repertoire has been introduced to produce human antibodies will not always result in high affinity antibodies probably due to a sub optimally functioning immune system of the transgenic animals. Im- munising immune compromised animals that have received human immune competent cells is another way of developing human antibodies, however the length of the immunisation strategy that can be used in this set-up, will not guaranty high affinity antibodies.

It is therefore also an object of the present invention to provide humanized antibod- ies directed against a target protein, where fully human antibodies generated in one of the above mentioned animal systems are combined with CDR sequences from high affinity anti- bodies generated in another animal, wherein both the fully human antibody and the CDR se- quences from high affinity antibodies generated in another animal are directed against the same target protein.

Thus, in a further aspect, the invention relates to an isolated humanized antibody, that immunoreacts with an epitope P present on a protein, wherein the human framework amino acid sequences of said humanized antibody are derived from a human antibody, wherein said human antibody immunoreacts with an epitope H present on said protein, said epitope H comprising an amino acid residue at a particular position of said protein, which amino acid residue at this particular position of said protein is also comprised within the epi-

tope P with which said humanized antibody immunoreacts. It is to be understood, that this protein may be any protein. The protein may be a therapeutic target protein.

The phrases"framework amino acid sequence"and"framework sequence" individually refer to an amino acid sequence of an antibody that is not a CDR sequence. A framework sequence that is derived from a parent antibody consists of the amino acids present in the corresponding non-CDR sequence of the parent antibody from which such sequences are considered to be derived. Unless otherwise stated or clearly contradicted by context, any framework sequence incorporated in any antibody described herein can be substituted with a suitable framework sequence variant. A framework sequence variant can be any sequence of amino acids that (a) has similar biological property as an antibody produced by a mammal in response to an antigen of interest (e. g. , permits immunogenic binding of the antibody of which it is a part to its target, such as TF, preferably with an affinity that is at least about as great, if not greater, than its mammalian-produced counterpart and desirably allows the antibody to retain a similar biological effect as its mammalian-produced counterpart-e. g. , a similar level of inhibition of TF binding to FVlla) and (b) has a high level of amino acid sequence identity with its mammalian-produced counterpart (e. g. , has at least about 80%, at least about 85%, at least about 90%, or at least about 95% identity to).

Typically, a framework sequence variant will vary by only a few amino acids from its mammalian-produced counterpart (e. g. , a variant may differ by 3,2, or 1 substitutions, deletions, and/or additions) and/or will vary only in terms of conservative amino acid substitutions. Often, a framework sequence variant will consist essentially of the same sequence as the framework sequence of its mammalian-produced counterpart.

In one embodiment of the invention the CDR amino acid sequences of the isolated humanized antibody are derived from a parent monoclonal antibody. In one embodiment the parent monoclonal antibody is a mouse monoclonal antibody. In one embodiment, the CDR amino acid sequences is variant CDR amino acid sequences relative to the CDR amino acid sequences of a parent monoclonal antibody. The terms"variant", as used herein with respect to CDR amino acid sequences, designates, designates CDR amino acid sequences, wherein one or more amino acids have been substituted by another amino acid and/or wherein one or more amino acids have been deleted and/or wherein one or more amino acids have been inserted into the CDR amino acid sequence. The"variant"within this definition still have CDR functionality. In one embodiment the variant CDR amino acid sequences is 70 % identical with the CDR sequences of the parent monoclonal antibody. In one embodiment the variant CDR amino acid sequences is 80 % identical with the CDR sequences of the parent mono- clonal antibody. In one embodiment the variant CDR amino acid sequences is 90 % identical with the CDR sequences of the parent monoclonal antibody. In one embodiment the variant

CDR amino acid sequences is 95 % identical with the CDR sequences of the parent mono- clonal antibody. In one embodiment the variant CDR amino acid sequences has one amino acid substitution relative to the CDR sequences of the parent monoclonal antibody. In one embodiment the variant CDR amino acid sequences has two amino acid substitutions rela- tive to the CDR sequences of the parent monoclonal antibody. In one embodiment the vari- ant CDR amino acid sequences has three amino acid substitutions relative to the CDR se- quences of the parent monoclonal antibody. In one embodiment the variant CDR amino acid sequences has four amino acid substitutions relative to the CDR sequences of the parent monoclonal antibody.

In one embodiment of the invention the isolated humanized antibody is a Fab frag- ment.

In a further embodiment of the invention the isolated humanized antibody is a F (ab) 2 fragment.

In a further embodiment of the invention the isolated humanized antibody is a F (ab') 2 fragment.

In a further embodiment of the invention the isolated humanized antibody is a single chain Fv fragment.

In a further embodiment of the invention the isolated humanized antibody has hu- man framework amino acid sequences derived from a human antibody, that immunoreacts with a second epitope present on human TF. In one embodiment this second epitope com- prises an amino acid residue at a particular position of human TF, which amino acid residue at this particular position of human TF is also comprised within the first epitope with which said humanized antibody immunoreacts. In one embodiment first epitope comprises an amino acid residue at a particular position of human TF selected from the list consisting of Trp45, Lys46 and Tyr94. In a further embodiment second epitope comprises an amino acid residue at a particular position of human TF selected from the list consisting of Trp45, Lys46 and Tyr94.

The terms"human antibody", "human antibodies", "human TF antibody", and"hu- man TF antibodies", as used herein, is intended to include antibodies having variable and constant regions derived from human germline immunoglobulin sequences. The human anti- bodies of the invention may include amino acid residues not encoded by human germline immunoglobulin sequences (e. g. , mutations introduced by random or site-specific mutagene- sis in vitro or by somatic mutation in vivo), for example in the CDRs and in particular CDR3.

However, the term"human antibody", as used herein, is not intended to include antibodies in which CDR sequences derived from the germline of another mammalian species, such as a

mouse, have been grafted onto human framework sequences, e. g. the so-called humanized antibodies.

In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF within the range of 10-15-10-8 M. It is to be understood, that the Kd for humanized antibody binding to human TF referred to is as determined in an assay, wherein the humanized antibody is immobilized (see assay 6).

In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF within the range of 10-'5-10-'° M.

In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF within the range of 10-15-10-12 M.

In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF lower than 10-8 M. In a further embodiment of the invention the iso- lated humanized antibody has a Kd for binding to human TF lower than 10-9 M In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF lower than 10-1° M. In a further embodiment of the invention the isolated humanized anti- body has a Kd for binding to human TF lower than 10-11 M. In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF lower than 10-12 M. In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF lower than 10-13 M. In a further embodiment of the invention the isolated humanized antibody has a Kd for binding to human TF lower than 10-14 M. In a further em- bodiment of the invention the isolated humanized antibody has a Kd for binding to human TF lower than 10-15 M.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a TF-induced clot assay and selecting humanized antibody which inhibit clot formation in this assay with an IC50 value lower than 1 nM. In a fur- ther embodiment the IC50 value is lower than 500 pM. In a further embodiment the IC50 value is lower than 200 pM. In a further embodiment the IC50 value is lower than 100 pM. In a fur- ther embodiment the ICgo value is lower than 50 pM. In a further embodiment the IC 50 value is lower than 10 pM. In a further embodiment the In50 value is lower than 5 pM.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a TF-induced clot assay and selecting humanized antibody which inhibit clot formation in this assay with an IC50 value lower than the IC50 value of FFR-rFVlla + 1 nM, such as lower than the IC ! 50 value of FFR-rFVlla + 500 pM, preferably lower than the IC5o value of FFR-rFVlla + 200 pM, preferably lower than the IC5o value of FFR-rFVlla + 100 pM, such as lower than the IC50value of FFR-rFVlla + 50 pM, preferably

lower than the IC50 value of FFR-rFVlla + 10 pM, more preferably lower than the IC50 value of FFR-rFVlla + 5 pM, more preferably lower than the Iso value of FFR-rFVI la.

It is to be understood, that the specific IC. 5o values referred to in the TF-induced clot assay is when using normal human plasma.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FXa generation assay and selecting humanized antibody which inhibit FXa generation with an ICSO value lower than 100 nM (in an assay with a FVlla concentration of 0.1 nM) In a further embodiment the IC5ovalue is lower than 10 nM.

In a further embodiment the IC50 value is lower than 5 nM. In a further embodiment the IC5c) value is lower than 1 nM. In a further embodiment the IC50 value is lower than 0.1 nM.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FXa generation assay and selecting humanized antibody which inhibit FXa generation with an IC50 value lower than the IC50 value of FFR- rFVlla + 100 nM (using 0.1 nM FVIIa in the assay), such as lower than the IC50 value of FFR- rFVlla + 10 nM, preferably lower than the ICSO value of FFR-rFVlla + 5 nM, preferably lower than the IC50 value of FFR-rFVlla + 1 nM, more preferably lower than the IC50 value of FFR- rFVlla + 0.1 nM, more preferably lower than the IC 50 value of FFR-rFVlla.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FVlla/TF amidolytic assay and selecting human- ized antibody which inhibit TF-induced FVlla amidolytic activity with an ICSO value lower than 100 nM (in an assay with a FVlla concentration of 10 nM). In a further embodiment the IC50 value is lower than 40 nM. In a further embodiment the IC50value is lower than 20 nM. In a further embodiment the IC50 value is lower than 10 nM. In a further embodiment the IC50 value is lower than 5 nM.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FVlla/TF amidolytic assay and selecting human- ized antibody which inhibit TF-induced FVI la amidolytic activity, with an IC50 value lower than the ICSO value of FFR-rFVlla + 100 nM (using 10 nM FVlla in the assay), such as lower than the IC50 value of FFR-rFVlla + 40 nM, preferably lower than the IC50 value of FFR-rFVlla + 20 nM, preferably lower than the IC50 value of FFR-rFVlla + 10 nM, more preferably lower than the IC50value of FFR-rFVI la.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FVlla competition assay and selecting humanized antibody which compete with FVIIa binding.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a TF ELISA assay comprising TF and selecting hu- manized antibody which bind human TF.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a direct TF ELISA assay comprising immobilized TF and selecting humanized antibodies which binds immobilized human TF.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in an indirect TF ELISA assay comprising TF in solu- tion and selecting humanized antibodies which binds human TF in solution In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FXa generation assay on TF expressing cell and selecting humanized antibodies which inhibits FXa generation on TF expressing cell with an IC50 value lower than 500 nM (in an assay with a FVlla concentration of 1 nM). In a further embodiment the IC50 value is lower than 100 nM. In a further embodiment the In50 value is lower than 50 nM In a further embodiment the IC50 value is lower than 10 nM In a further em- bodiment the IC50 value is lower than 5 nM.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a FXa generation assay on TF expressing cell and selecting humanized antibodies which inhibits FXa generation on TF expressing cell with an IC. 50 value lower than the In50 value of FFR-rFVlla + 500 nM (using 1 nM FVI la in the assay), such as lower than the Ices value of FFR-rFVlla + 100 nM, preferably lower than the ICgo value of FFR-rFVlla + 50 nM preferably lower than the iCso vaiue of FFR-rFVlla + 10 nM, more preferably lower than the ICgo value of FFR-rFVlla + 5 nM, more preferably lower than the Iso value of FFR-rFVI la.

The term"TF expressing cell"mean any mammalian cell, that expresses human TF.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a whole cell TF binding assay and selecting human- ized antibodies which competes with FVlla binding to human TF expressed on the cell sur- face of whole cells.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a biosensor assay and selecting humanized anti- bodies with a Kd value for binding to human TF lower than 100 nM. In a further embodiment the Kd value for binding to human TF is lower than 10 nM. In a further embodiment the Kd value for binding to human TF is lower than 5 nM. In a further embodiment the Kd value for binding to human TF is lower than 1 nM. In a further embodiment the Kd value for binding to human TF is lower than 0.5 nM. In a further embodiment the Kd value for binding to human

TF is lower than 10-1° M. In a further embodiment the Kd value for binding to human TF is lower than 10-11 M. In a further embodiment the Kdvalue for binding to human TF is lower than 10-12 M. In a further embodiment the Kdvalue for binding to human TF is lower than 10-13 M. In a further embodiment the Kd value for binding to human TF is lower than 10-14 M. In a further embodiment the Kd value for binding to human TF is lower than 10-15 M.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a MAPK signalling assay and selecting humanized antibodies which inhibits FVlla-induced activation of the MAPK signalling. In one embodiment the humanized antibody inhibits FVlla-induced activation of the MAPK signalling with 98 %. In one embodiment the humanized antibody inhibits FVlla-induced activation of the MAPK sig- nalling with 90 %. In one embodiment the humanized antibody inhibits FVlla-induced activa- tion of the MAPK signalling with 70 %. In one embodiment the humanized antibody inhibits FVlla-induced activation of the MAPK signalling with 50 %. In one embodiment the humanized antibody inhibits FVlla-induced activation of the MAPK signalling with 30 %.

The term"MAPK signalling"means a cascade of intracellular events that mediate activation of Mitogen-Activated-Protein-Kinase (MAPK) or homologues thereof in response to various extracellular stimuli. Three distinct groups of MAP kinases have been identified in mammalian cells : 1) extracellular-regulated kinase (Erk1/2 or p44/42), 2) c-Jun N-terminal kinase (JNK) and 3) p38 kinase. The Erk1/2 pathway involves phosphorylation of Erk 1 (p 44) and/or Erk 2 (p 42). Activated MAP kinases e. g. p44/42 MAPK can translocate to the nu- cleus where they can phosphorylate and activate transcription factors including (Elk 1) and signal transducers and activators of transcription (Stat). Erk1/2 can also phosphorylate the kinase p90RSK in the cytoplasm or in the nucleus, and p90RSK then can activate several transcription factors.

The term"protein kinase"is intended to indicate an enzyme that is capable of phosphorylating serine and/or threonine and/or tyrosine in peptides and/or proteins.

The term"FVlla-induced activation of the MAPK signalling"is intended to indicate that FVlla binds to TF in a mammalian cell and thereby induce MAPK signalling.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a gene expression analysis assay (assay 15) and selecting humanized antibodies which inhibits FVlla induced up-regulation of genes inde- pendently selected from the list comprising Fra-1, Id2, and Cyr61.

It is to be understood, that antibodies against TF, which inhibits the activity of TF may bind different epitopes present on TF and may inhibit both the binding of FVlla or the binding of FX or FXa to human TF. It is an object of the present invention to select antibod-

ies, which inhibits the binding of FVlla to human TF and thereby the FVlla induced intracellu- lar signalling.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in a human cancer assay (assay 13) and selecting humanized antibodies which inhibits growth or metastasis of human cancers.

In a further embodiment of the invention the isolated humanized antibody inhibits FVlla induced up-regulation of genes independently selected from the list comprising Fra-1, Id2, and Cyr61.

In a further embodiment of the invention the isolated humanized antibody does not inhibit binding of FX or FXa to human TF.

In a further embodiment of the invention the isolated humanized antibody inhibits the intracellular activity of TF.

In a further embodiment of the invention the method for preparation of a humanized antibody comprises testing antibodies in an epitope mapping assay and selecting humanized antibodies which binds preferred epitopes on TF. In one embodiment the preferred epitope comprises one or more of the residues Trp45, Lys46 and Tyr94. In one embodiment the pre- ferred epitope comprises the residue Trp45. In one embodiment the preferred epitope com- prises the residue Lys46. In one embodiment the preferred epitope comprises the residue Tyr94.

In a further embodiment of the invention the isolated humanized antibody binds to an epitope within the interface between TF and FVlla.

The residues in TF that are responsible for the interaction between the protease domain of FVlla and TF determined from the X-ray structure (Banner et al. 1996 Nature, 380: 41-46) are; Ser39, Gly43, Trp45, Ser47, Phe50, Arg74, Phe76, Tyr94, Pro92. This interface between the protease domain of FVlla and TF is characterized as being a complex interface region containing many intermolecular hydrogen bonds allowing many fine contacts between TF and FVlla to obtain high specificity in binding process of FVlla.

The present invention also relates to high affinity humanized antibodies to TF. The TF surface containing the contact interface for the protease domain of FVlla holds a specific topology that are prone to react to create protein-protein interactions, wherein another type of protein-protein interaction is the complex formation between an antibody and a protein ligand.

One aspect of the present invention is high affinity humanized antibodies, that are immunoreacting with the contact interface for the protease domain of FVlla.

Humanized TF antibodies of the present invention act as antagonists for TF-mediated induction of coagulation, thus inhibiting the binding of coagulation FVlla to TF

and thereby blocking the production of thrombin and the subsequent deposition of fibrin.

Humanized TF antibodies are particularly useful for administration to humans to treat a vari- ety of conditions involving intravascular coagulation. As such, humanized TF antibodies may be useful for inhibiting TF activity resulting in, for example, the inhibition of blood coagulation, thrombosis or platelet deposition. Furthermore, humanized TF antibodies according to the present invention, which acts to inhibit the cellular functions of TF, the signalling function of TF, may be useful in conditions like sepsis, inflammation, atherosclerosis, restenosis, or can- cer.

Humanized TF antibodies may be useful in a variety of diseases. Included are thrombotic or coagulopathic related diseases or disorders including inflammatory response and chronic thromboembolic diseases or disorders associated with fibrin formation including vascular disorders such as deep venous thrombosis, arterial thrombosis, post surgical thrombosis, coronary artery bypass graft (CABG), percutaneous transdermal coronary angioplastry (PTCA), stroke, tumor growth, tumor metastasis, angiogenesis-related disease (s), thrombolysis, atherosclerosis, arteriosclerosis e. g. arteriosclerosis following angioplastry, and restenosis e. g. restenosis following angioplastry, acute and chronic indications such as inflammation, septic chock, septicemia, hypotension, adult respiratory distress syndrome (ARDS), systemic inflammatory response syndrome (SIRS), disseminated intravascular coagulopathy (DIC), pulmonary embolism, pathological platelet deposition, myocardial infarction, or the prophylactic treatment of mammals with atherosclerotic vessels at risk for thrombosis, venoocclusive disease following peripheral blood progenitor cell (PBPC) transplantation, hemolytic uremic syndrome (HUS), thrombotic thrombocytopenic purpura (TTP) and other diseases or disorders. The humanized TF antibodies may be used to prevent the occurrence of thromboembolic complications in identified high risk patients, such as those undergoing surgery or those with congestive heart failure. The humanized TF antibodies may be particularly useful in the treatment of intimal hyperplasia or restenosis due to acute vascular injury. Acute vascular injuries are those which occur rapidly (i. e. over hours to months, in contrast to chronic vascular injuries (e. g. atherosclerosis) which develop over a lifetime. Acute vascular injuries often result from surgical procedures such as vascular reconstruction, wherein the techniques of angioplasty, endarterectomy, atherectomy, vascular graft emplacement or the like are employed.

Hyperplasia may also occur as a delayed response in response to, e. g. , graft emplacement or organ transplantation. Since humanized TF antibodies is more selective than heparin, binding only TF which has been exposed at sites of injury, and because humanized TF antibodies does not destroy or inhibit other coagulation proteins, it will be more effective and

less likely to cause bleeding complications than heparin when used prophylactically for the prevention of deep vein thrombosis.

As shown in the examples that follow, the humanized TF antibodies of the present invention is able to bind selectively to cell-surface TF and inhibit its functional activity by inhibiting the binding of coagulation FVlla to TF. Humanized TF antibodies which maintain binding to TF inhibit platelet accumulation at the site of vascular injury by blocking the production of thrombin and the subsequent deposition of fibrin.

Due to the ability of humanized TF antibodies to block thrombin generation and limit platelet deposition at sites of acute vascular injury, humanized TF antibodies which maintain binding to TF thereby inhibiting FVlla binding can be used to inhibit vascular restenosis.

Compositions comprising humanized TF antibodies are particularly useful in methods for treating patients when formulated into pharmaceutical compositions, where they may be given to individuals suffering from a variety of disease states to treat coagulation- related conditions. Such humanized TF antibodies, capable of binding TF and inhibiting FVlla binding to TF, may possess a longer plasma half-life and thus a correspondingly longer period of anticoagulant activity when compared to other anticoagulants. Among the medical indications for the subject compositions are those commonly treated with anticoagulants, such as, for example, deep vein thrombosis, pulmonary embolism, stroke, disseminated intravascular coagulation (DIC), fibrin deposition in lungs and kidneys associated with sepsis, antiphospholipid syndrome (APS), atherosclerosis and myocardial infarction. The compositions can be used to inhibit vascular restenosis as occurs following mechanical vascular injury, such as injury caused by surgery, microsurgery, balloon angioplasty, endarterectomy, reductive atherectomy, stent placement, laser therapy or rotablation, or as occurs secondary to vascular grafts, stents, bypass grafts or organ transplants. The compositions can thus be used to inhibit platelet deposition and associated disorders. Thus, a method of inhibiting coagulation, vascular restenosis or platelet deposition, for example, comprises administering to a patient a composition comprising humanized TF antibodies in an amount sufficient to effectively inhibit coagulation, vascular restenosis or platelet deposition. The methods also find use in the treatment of acute closure of a coronary artery in an individual (e. g. acute myocardial infarction), which comprises administering the humanized TF antibodies, in conjunction with tissue plasminogen activator or streptokinase, and can accelerate tPA induced thrombolysis. The humanized TF antibodies are given prior to, in conjunction with, or shortly following administration of a thrombolytic agent, such as tissue plasminogen activator.

According to the invention, human monoclonal antibodies directed against human TF may be produced by immunizing transgenic mice (Obtained from Medarex) carrying parts

of the human immune system rather than the mouse system with human TF. Splenocytes from these transgenic mice are used to produce hybridomas that secrete human monoclonal antibodies as described (see, e. g. , Wood et al. International Application WO 91/00906, Kucherlapati et al. PCT publication WO 91/10741; Lonberg et al. International Application WO 92/03918; Kay et al. International Application 92/03917; Lonberg, N. et al. 1994 Nature 368: 856-859; Green, L. L. et al. 1994 Nature Genet. 7: 13-21; Morrison, S. L. et al. 1994 Proc. Natl. Acad. Sci. USA 81: 6851-6855; Bruggeman et al. 1993 Year Immunol 7: 33-40; Tuaillon et al. 1993 PNAS 90: 3720-3724; Bruggeman et al. 1991 Eur J Immunol 21: 1323- 1326).

Human monoclonal antibodies directed against human TF may also be produced by phage display. Human antibody libraries can be constructed from immunized persons and displayed on the surface of filamentous phage. High-affinity human single-chain Fv (ScFv) and Fab antibody fragments have in numerous of cases been isolated from such libraries using a panning technique in which the antigen of interest is immobilised on a solid surface i. e. microtiter plates or beads (Barbas C. F., III and Burton, D. R. Trends. Biotechnol. 1996, 14: 230-234; Aujame L. et al, Hum. Antibodies 1997,8 : 155-68). Phage display of large naive libraries has also made it possible to isolate human antibodies directly without immunization (DeHaard H. J. et al J. Biol. Chem. 1999,18218-18230).

BRIEF DESCRIPTION OF THE DRAWINGS The present invention is described in further detail in the examples with reference to the appended drawings wherein Figure 1. Schematic presentation of an examplified screening assay for selection of human monoclonal high affinity antibodies against TF.

Figure 2. Detailed schematic representation of screening assays 1-3 as described in exam- ple 1.

Figure 3. Detailed schematic representation of screening assays 4-7 as described in exam- ple 1.

Figure 4. Detailed schematic representation of screening assays 8-10 as described in exam- ple 1.

Figure 5. An example of screening antibodies by assay no. 4. Inhibition of sTF/FVlla amido- lytic activity by FFR-rFVlla (closed circles) and the human anti TF monoclonal antibody HuTF-31 F2 (open circles).

Figure 6. An example of screening antibodies by assay no. 5. Inhibition of factor Xa genera- tion by FFR-rFVlla (closed circles) and the human anti-TF monoclonal antibody HuTF-31 F2 (open circles).

Figure 7. An example of screening antibodies by assay no. 7. Inhibition of human TF-induced clotting by FFR-rFVlla (closed circles) and the human anti TF monoclonal antibody HuTF- 31 F2 (open circles).

Figure 8. An example of screening antibodies by assay no. 10. Only anti-TF monoclonal an- tibodies preventing FVlla binding inhibits TF/FVlla-mediated signaling.

Figure 9. Human anti TF Mab inhibit FVlla induced phosphorylation of p44/42 MAPK (assay no. 10). BHK cell transfected with TF were serum-starved for 2 hr to make cells quiescent.

The antibodies HuMab 30F5 (500 nM) and HuMab 31 F2 (500 nM) were added to the cells 15 min prior addition of FVlla (15 nM). Cells were lysed and proteins were separate on SDS- PAGE and transferred to nitrocellulose by electroblotting. Western blot analysis was per- formed using polyclonal phospho-specific antibodies to p44/42 MAPK. Secondary antibodies were anti-rabbit IgG conjugated to Horse Radish Peroxidase. Detection of chemilumines- cence was performed using a cooled CCD-camera. The bands on the digitalized picture were quantified and the band obtained with FVlla was set to 100%. When cells were pre-incubated with HuMab 30F5 (500 nM) a 50% reduction in the phosphorylated band was observed and when cells were pre-incubated with HuMab 31 F2 (500 nM) a 25 % reduction was observed.

In conclusion, this experiment show that the human antibodies against TF (30F5 and 31 F2) partially inhibited the FVlla induced phosphorylation of p44/42 MAPK. Similar results were obtained using 50 nM FVlla.

Figure 10. An example of screening antibodies by assay no. 16. The figure demonstrates the inhibition of TF intracellular activity in TF expressing cells by monoclonal antibodies against TF. Anti TF Mab B inhibits TF intracellular activity, while Anti-TF Mab A do not.

Figure 11. An example of screening antibodies by assay no. 12. Velocity profile of thromboe- lastograms obtained with 0.5 nM of FFR-rFVlla and the human anti-TF antibody HuTF-31 F2.

The present invention is further illustrated by the following examples which, however, are not to be construed as limiting the scope of protection. The features disclosed in the forego- ing description and in the following examples may, both separately and in any combination thereof, be material for realizing the invention in diverse forms thereof.

EXAMPLES Example 1 Preparation of human monoclonal antibodies immunospecific for human TF.

Reagents.

Human TF can be isolated from human brain as described by Rao, L. V. M. , Throm- bosis Research, 51: 373-384 1988.

Lipidated recombinant human TF (Dade Innovin, Baxter) can also be used as hu- man thromboplastin reagent. Rat, rabbit, baboon, and pig thromboplastin are prepared from brain tissue. Two volumes of 45°C 0.9 % NaCI are added to the brain tissue, and the tissue is homogenized with a manual glass homogenisator. After 30 min incubation at 45°C with oc- casional shaking, the samples are centrifuged 20 min at 2000 x g. The precipitate are dis- carded, and the supernatant is aliquote and stored at-80°C until use.

Relipidated TF may be obtained by reconstitution of recombinant human full length TF (American Diagnostica #4500) into phospholipid vesicles (PC/PS 75/25).

Biotinylated human TF is produced as follows : Biotin-NHS (n-succinimido biotin, Sigma H-1759) is dissolved in DMF (dimethylformamid) at a concentration of 1.7 mg/ml. 1 mg/ml of human TF in 0.1 M NaHC03 buffer is added 60 gl of the biotin-NHS solution and is allowed to react for 4 hours at room temperature. The reacting solution containing the bioti- nylated TF is dialysed against PBS-buffer over night.

FVlla used is recombinant human FVlla prepared as described by Thim, L. et al. Biochem 27: 7785-7793,1988. sTF: Recombinant human soluble TF,-209 expressed in E. coli and purified essentially as de- scribed by Freskgård, P. -O. et al. Protein Sci. 5,1531-1540 (1996).

S2288: reconstituted in H20 to 17. 24 mg/ml (Chromogenix) FX: Purified human plasma FX (HFX, Enzyme Research Laboratories Ltd.)

FXa: Purified human plasma FX activated with Russel's Viper Venom (HFXa, Enzyme Re- search Laboratories Ltd.) Chromozyme X: Dissolved in H20 to 1.25 mg/ml. (Boehringer Mannheim) 51-FVlla is obtained by standard radiolabelling procedures.

FFR-rFVlla : FVlla blocked in the active site with D-Phe-L-Phe-L-Arg-chloromethyl ketone.

Prepared as described by Sorensen B. B. et al. J. BioLChem. 272: 11863-11868,1997.

Immunization.

Human TF is emulsified in Freunds Complete Adjuvant. HuMab mice or hybrids therof (Medarex) are given 40 Hg by a subcutaneous injection. After 14 and 28 days, and eventually more times with intervals of 14 days the mice are boosted with a similar injection of 20 jug of TF in Incomplete Freunds Adjuvant. Ten days after the last injection a blood sample is taken and sera are tested for human TF specific antibodies by TF ELISA (Assay 1 and 2).

Fusion.

Mice with positive serum test from assay 1-3 are boosted with 20 jug of human TF by an intravenous injection and sacrificed after three days. The spleen is removed aseptically and dispersed to a single cell suspension.

Fox-myeloma cells are grown in CD Hybridoma medium (Gibco 11279-023).

Fusion of spleen cells and myeloma cells (P3x63 Ag8.653, ATCC CRL-1580), and the Sp2/0 (ATCC CRL-1581) myeloma cell lines for our fusions are done by the PEG- methods (Kohler, G & Milstein C. (1976), European J. Immunology, 6: 511-19). Cells are seeded in microtiter plates and incubated at 37 °C. Medium is changed three times over the next two weeks. 100 gel of supernatant from hybridoma cells is removed from each well and tested for TF specific antibodies in TF ELISA (Assay 1 and 2).

Example 2: Screening.

The various assays used in the screening of serum and culture supernatants for se- lecting specific humanized or human antibodies are described in the following : Direct TF ELISA assay (Assay 1): Nunc immunoplates are coated overnight at 4°C with 1 pg/ml of human sTF in PBS.

Plates are blocked with blocking buffer (TBS with 5 mM Caca2 an 2% BSA) and washed in

TBS+ 0.05 % Tween 20, and the supernatants from the host cells are added. After incubation at room temperature for 1 hour, plates are washed and anti-human IgG labelled with horse- radish peroxidase (HRPO) is added. After another hour of incubation, plates are washed and developed with TMB-substate (Kem-EN-Tec) as described by the manufactures. Absorbance at 450 nm is measured on an ELISA-reader.

Indirect TF ELISA assay (Assay 2): Nunc-immunoplates are coated with 0.5 pg/ml of goat anti-human IgG (Southern Biotechnology Associates, Cat-#2040-1) in PBS and incubated overnight at 4°C. Plates are blocked with blocking buffer (TBS with 5 mM Caca an 2% BSA) and washed in TBS+ 5 mM CaCI2 + 0. 05 % Tween 20. Culture supernatants from the host cells are added and the plates incubated for 1 hour at room temperature. After another wash, biotinylated human sTF are added at a concentration of 1 ug/ml, and incubated for 1 hour. After washing, 100 RI of a Streptavidin-HRPO solution is added and incubated for 1 hour. Plates are developed with TMB-substate as described for assay 1.

FVlla competition assay (Assay 3): Nunc-immunoplates are incubated with human sTF (conc 5 pg/ml in PBS) over night, 4 °C.

Plates are washed and blocked in TBS buffer with 5 mM CaCl2 and 2 % BSA. Anti-human TF antibodies are added and plates are incubation for 2 hours. Plates are washed before bioti- nylated human FVlla are added (1 lig/ml in TBS buffer with 5 mM CaCI2 and 2 % BSA) and the plates incubated for 1 hour. Plates are washed before addition of HRPO-labeled Strepta- vidin and incubated for 45 min. Plates are washed again before development with TMB sub- strate (Kem-EN-Tec) as described by the manufactures.

Inhibition of FVlla/sTF amidolytic activity (Assay 4): Inhibition of FVlla-TF catalyzed amidolytic activity by anti-human TF antibodies is tested employing soluble human TF (10 nM), recombinant human FVlla (10 nM) and increas- ing concentrations of antibodies (0.0122-50 nM). Varying concentrations of anti-human TF antibodies or FFR-rFVlla are preincubated with 10 nM sTF and 10 nM FVlla in BSA buffer (50 mM Hepes, pH 7.4, 100 mM NaCI, 5 mM CaCl2 and 1 mg/ml BSA) for 60 min at room temperature before addition of substrate S2288 (1.2 mM, Chromogenix). The colour devel- opment is measured continuously for 30 min at 405 nm. Amidolytic activity is presented as mOD/min. IC50 values for inhibition of FVlla/TF amidolytic activity by the antibodies may be calculated. The IC5o value for FFR-rFVlla is 7 +/-3 nM in this assay.

Inhibition of FXa generation (Assay 5).

Lipidated TF (10 pM), FVlla (100 pM) and anti-TF antibodies or FFR-rFVlla (0-50 nM) in BSA buffer (see assay 4) are incubated 60 min at room temperature before FX (50 nM ) is added. The reaction is stopped after another 10 min by addition of 1/2 volume stopping buffer (50 mM Hepes, pH 7.4, 100 mM NaCI, 20 mM EDTA). The amount of FXa generated is determined by adding substrate S2765 (0.6 mM, Chromogenix, and measuring absorb- ance at 405 nm continuously for 10 min. iCso values for antibody inhibition of FVlla/lipidated TF-mediated activation of FX may be calculated. The ICSO value for FFR-rFVlla is 51 +/-26 pM in this assay.

Biosensor assay (Assay 6): antibodies are tested on the Biacore instrument by passing a standard solution of anti-human TF antibody over a chip with immobilized antibody to human IgG. This is followed by different concentrations of sTF in 10 mM hepes pH 7.4 containing 150 mM NaCI, 10 mM CaCl2 and 0.0003 % polysorbate 20. Kd values are calculated from the sensorgrams using the integrated Biacore evaluation software.

TF-dependent clotting assay (Assay 7): The assay is carried out on an ACL300 Research clotting apparatus (ILS Laborato- ries). Dilutions of anti-human TF antibodies in 50 mM imidazole, pH 7.4, 100 mM NaCI, 0.1 % BSA are mixed with 25 mM Caca in the ratio of 2 to 5 and added to sample cups in the clotting apparatus. Thromboplastin from human, rat, rabbit, baboon, or pig diluted with the imidazole buffer to give clotting time of approximately 30 sec in samples without antibody is placed in reagent reservoir 2, and human, rat, rabbit, baboon, or pig plasma, in reagent res- ervoir 3. During the analysis 70 pl of the antibody and CaCl2 mixture is transferred to 25 pl thromboplastin reagent and preincubated 900 sec before addition of 60 pl plasma and meas- uring of the clotting time. Maximal clotting time is set to 400 sec. A dilution of the throm- boplastin is used as standard curve for converting clotting times into TF activity relative to the control without anti-TF antibodies or FFR-rFVlla added. The tCeo value for FFR-rFVlla is 4.4 +/-0. 4 pM in this assay.

Inhibition of FVlla/cell surface TF catalyzed activation of FX by anti TF antibodies (As- say 8): Monolayers of cells expressing human TF, e. g. human lung fibroblasts WI-38 (ATTC No. CCL-75), human bladder carcinoma cell line J82 (ATTC No. HTB-1), human keratinocyte cell line CCD 1102KerTr (ATCC no. CRL-2310), human glioblastoma cell line U87, or human breast cancer cell line MDA-MB231, are employed as TF source in FVlla/TF catalyzed acti- vation of FX. Confluent cell monolayers in a 24-, 48-or 96-well plate are washed one time in buffer A (10 mM Hepes, pH 7.45, 150 mM NaCI, 4 mM KCI, and 11 mM glucose) and one time in buffer B (buffer A supplemented with 1 mg/ml BSA and 5 mM Ca2+). FVlla (1 nM), FX (135 nM) and varying concentrations of antibody (or FFR-rFVlla) in buffer B are simultane- ously added to the cells. Alternatively the cells are preincubated 15 min with anti-TF antibod- ies or FFR-rFVlla before addition of rFVlla and FX. FXa formation is allowed for 15 min at 37°C. 50-, ul aliquots are removed from each well and added to 50 pl stopping buffer (Buffer A supplemented with 10 mM EDTA and 1 mg/ml BSA). The amount of FXa generated is de- termined by transferring 50 pi of the above mixture to a microtiter plate well and adding 25 ut Chromozym X (final concentration 0.6 mM) to the wells. The absorbance at 405 nm is meas- ured continuously and the initial rates of colour development are converted to FXa concen- trations using a FXa standard curve. The ICso value for FFR-rFVlla-is 1.5 nM in this assay.

Inhibition of 1251-FVlla binding to cell surface TF by antibody (Assay 9): Binding studies are employed using cells expressing human TF, e. g. human lung fi- broblasts WI-38 (ATTC No. CCL-75), human bladder carcinoma cell line J82 (ATTC No.

HTB-1), human keratinocyte cell line CCD 1102KerTr (ATCC no. CRL-2310), human glioblastoma cell line U87, or human breast cancer cell line MDA-MB231. Confluent monolayers in 24-well tissue culture plates are washed once with buffer A (see assay 8) and once with buffer B (see assay 8). The monolayers are preincubated 2 min with 100 pl cold buffer B. Varying concentrations of antibodies (or FFR-rFVlla) and radiolabelled FVlla (0.5 nM 1251-FVlla) are simultaneously added to the cells (final volume 200 pl). The plates are in- cubated for 2 hours at 4 °C. At the end of the incubation, the unbound material is removed, the cells are washed 4 times with ice-cold buffer B and lysed with 300 pi lysis buffer (200 mM NaOH, 1 % SDS and 10 mM EDTA). Radioactivity is measured in a gamma counter (Cobra, Packard Instruments). The binding data are analyzed and curve fitted using GraFit4 (Erithacus Software, Ltd. , (U. K.). The IC50 value for FFR-rFVlla is 4 nM in this assay.

Inhibition of FVlla/TF-induced p44/42 MAPK activation by anti TF antibody (Assay 10): The amount of phosphorylated p44/42 MAPK and/or Akt, and/or p90RSK is deter- mined by quantitative detection of chemiluminescence (Fujifilm LAS-1000) from western blot analysis. Cells expressing human TF, e. g. CCD1102KerTr, NHEK P166, human glioblastoma cell line U87, or human breast cancer cell line MDA-MB231, are cultured in medium with 0- 0.1 % FCS for 24 or 48 hours prior to the experiment to make cells quiescent. At the day of the experiment the cells must be 70-80% confluent. The experiment is performed by prein- cubating the cells with excess antibody or FFR-rFVlla in medium without serum for 30 min at 37°C before addition of 10-100 nM FVlla and incubating for 10 min. As a positive control of cell signaling, cells are treated with 10 % FCS for 10 minutes. Cells are washed 2 times in ice-cold PBS before cells are lysed in lysis buffer (20 mM Tris, 0. 1% Triton X-100,1 mM EDTA, 1 mM EGTA, 50 mM sodium-fluoride, 10 mM sodium ß-glycerophosphate, 5 mM so- dium pyrophosphate, 150 mM NaCI, pH 7.5 containing 0.1 mM 4- (2-aminoethyl) benzene- sulfonyl fluoride (AEBSF) and 1 mM benzamidine. Added just before use: 1 mM sodium or- thovanadate, 5 pg/ml leupeptin, 10 pg/ml aprotinin). Lysates were mixed with SDS-sample buffer and loaded on a SDS-polyacrylamide gel. A standard biotinylated protein marker is loaded on each gel. Proteins separated on the SDS-polyacrylamide gel were transferred to nitrocellulose by electroblotting, and the kinases p44/42 MAPK, Akt and p90RSK were visu- alized by immunoblotting with phosphospecific antibodies, and chemiluminiscence is quaniti- ated by Fujifilm LAS1000.

Epitope mapping assay (Assay 11): Preparation of soluble TF (sTF) variants. sTF variants (122C, W45C, K46C, Y94C, F140C, W158C, K201C) are constructed using in- verse PCR (QuikChange, Stratagene, La Jolla, CA, USA) using the wild type plasmid (Freskgard et al. Protein Sci. 5,1531-1540, 1996) as template. The wild type and variants are expressed and purified in E. coli as described elsewhere (Freskgard et al., Protein Sci. 5, 1531-1540, 1996) with some modifications. The cell lysis is performed by the X-press (Biox, Sweden) technique in 10 mM Tris-HCI buffer, pH 7.5 and thereafter resuspended in the same buffer with the addition of 1 mg of DNAse. The solution is centrifuged at 11000xg for 20 min at 4°C, and the inclusion bodies are denatured in 75 ml of 6 M GuHCI, 0.5 M NaCI, 20 mM Tris-HCI, pH 8.0. Refolding is achieved after 1-hour incubation at room temperature by dropwise diluting the denatured protein into a 1 L solution containing 50 mM Tris-HCI, 0.25 M NaCI, pH 8.0 with gentle stirring for approximately 2 hours. Purification is performed using Q- Sepharose Fast Flow (Pharmacia, Uppsala, Sweden) and FVlla affinity chromatography as described by Freskgard et al. (1996). The homogeneity of the protein is verified by SDS-

PAGE. The concentration is measured at A280nm and determined using a calculated extinction coefficient of 37440 M-1cm-1 (Gill and von Hippel, 1989).

MaxiSorp plates (Nunc-Immuno) are coated with wild type sTF and the variants (10 jug/mi) in TBS and blocked with blocking buffer (TBS with 0. 1% Tween 20 and 0.5% BSA).

The plates are washed with washing buffer (TBS and 0. 1% Tween 20). The anti-human TF antibodies are applied at a concentration of 1 ng/ml in blocking buffer and incubated for one hour. The plates are then washed (6x) using the washing buffer. The antibody binding is subsequently detected using an HRP-labelled anti-human IgG (Helica Biosystems, Inc) at a 1: 2000 dilution in blocking buffer using the TMBplus-substrate (Kem Tech Cat. 4390A). The final ELISA signal (OD450-620) is used as a measure of the binding of each antibody to all sTF variants.

Thromboelastography (Assay 12) Human thromboplastin (e. g. Innovin, Dade Bering, final dilution 50, 000 x) is mixed with CaC12 (final concentration 16.7 mM) and anti-TF antibodies and incubated 15 min at room temperature. Citrate-stabilized human whole blood (280 nui) is added to RoTEG sample cups (Pentapharm) and preheated 5 min at 37°C, before addition of 40 pl throm- boplastin/CaCI2/anti-TF antibody mixture. Thromboelastography is followed for one hour in a RoTEG apparatus (Pentapharm). Velocity profiles are obtained from the thrombograms using CoagPro SoftwareT"" (MedScience, Arhus, Denmark).

Example 3.

Human cancer assay. Investigating the effects of treatment with human anti-TF anti- bodies on growth and metastasis of human cancers in mouse models (Assay 13) Treatment: Human anti-TF antibodies given by bolus injection i. v.; 10 mg/kg = 0. 1mg/10g ; Injection- volume is 0.1 ml per 10 g mouse of either of three treatment solutions: A. Vehicle control

B. 1 mg/ml Human FFR-rFVlla C. 1 mg/ml anti-TF antibodies Description of models : 1. Primary growth and liver mefastasis of colon cancer Healthy female athymic mice (nulnu) aged 7-8 weeks are used. To destroy the re- sidual immunoresistance of the nude mice to the human cell implantation, the mice are rou- tinely irradiated at 5 Gy 2 days before human tumor grafting (Vogel et aL, 1997). Mice are challenged by tumor grafting of LS174T humancolon carcinoma cells (ATCC CCL 188) cul- tured in RPMI 1640 with 15% fetal calf serum (FCS) as described (Li et al., Human Gene Therapy 10: 3045-3053, 1999). In brief, the cells are harvested with trypsin-EDTA, washed twice, and resuspended in serumfree RPMI supplemented with sodium-heparinate solution (1 U/ml). A small left subcostal incision is then carried out in mice under anesthesia and 3 3 106 LS174T cells in 50 m lof phosphate-buffer ed saline (PBS) are injected into the spleen.

After 3 to 5 min, the spleen vessels are ligated and the spleen is surgically removed. This procedure will lead to a stable incidence of liver metastasis (more than 95%). The treatment with anti-TF antibodies will be initiated immediately after implantation and will last for the re- maining study period. On days 15 and 30 after tumor cell inoculation mice are sacrificed, the livers ae removed and weighed, and the number of visible tumor nodules on the liver sur- faces are counted. Liver samples are fixed overnight in AFA (5% acetic acid, 75% ethyl alco- hol, 2% formalin, 18% water), transferred to 100% ethanol, and embedded in paraffin, and 5- m m sections are prepared for histological quantification of metastatic nodules, for immuno- histochemistry and apoptosis quantification.

Study 1-1 : Aim: To examine the effect on macroscopical growth and liver metastasis of LS174T colon tumors in nude mice of anti-TF antibodies given bolus injection i. v.; 10 mg/kg.

Mice: 60 homozygotous nu/nu 6 weeks old NMRI males.

Groups Mice are randomly allocated in four groups of 15 and treated with solutions A, B, or C.

Termination: At a weight loss of > 20% or other objective signs of severe toxicity the animal is terminated.

Parameters: Tumor size in two orthogonal diameters is recorded daily during the growth phase. The body weight is recorded initially and 2-3 times per week. Postmor- tem determination of metastasis formation in the liver.

II. Primary growth and lung metastasis of mammary cancer Human breast carcinoma cells MDA-MB-231 (ATCC HTB26) are cultured in Dul- becco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS).

MDA-MB-231 cells (3 3 106) are injected subcutaneously in nude mice (7-to 8-week-old fe- male mice). Primary tumor growth and metastasis is evaluated as described previously (Li et al., Human Gene Therapy 12: 515-526, 2001) Study II-1 : Aim: To examine the effect on macroscopical growth and lung metastasis of MDA- MB-231 mammary tumors in nude mice of anti-TF antibodies given bolus in- jection i. v.; 10 mg/kg.

Mice: 60 homozygotous nu/nu 6 weeks old NMRI males.

Groups: Mice are randomly allocated in four groups of 15 and treated with solutions A, B, or C.

Termination: At a weight loss of > 20% or other objective signs of severe toxicity the animal is terminated.

Parameters: Tumor size in two orthogonal diameters is recorded daily during the growth phase. The body weight is recorded initially and 2-3 times per week. Postmor- tem determination of metastasis formation in the lung.

Ill. Primarygrowth of glioma tumorxenograffs The tumor cell line MG U373 is a human glioblastoma multiforme cell line, with high angiogenic activity, high vascular density and fast growth in nude mice. Tumors are inocu- lated in the flanks, following standard procedures (see enclosed protocols for experimental plan). The mice are observed twice daily for signs of toxicity and the tumors are measured daily in two perpendicular diameters.

Tumors are transplanted to the flanks of nu/nu homozygotous nude mice of NMRI background. The mice are 7-week-old males obtained from M&B (Ry, Denmark). Animals are kept in a gnotobiotic environment and they receives sterile food pellets and drinking water ad libitum.

Three different studies is conducted with the glioma tumor model : Study 111-1 : Aim: To examine the effect on macroscopical growth of U373 tumors in nude mice of anti-TF antibodies given bolus injection i. v.; 10 mg/kg.

Mice: 60 homozygotous nu/nu 6 weeks old NMRI males.

Groups: Mice are randomly allocated in four groups of 15 and treated with solutions A, B, or C.

Termination: At a weight loss of > 20% or other objective signs of severe toxicity the animal is terminated.

Parameters: Tumor size in two orthogonal diameters is recorded daily during the growth phase. The body weight is recorded initially and 2-3 times per week.

Study 111-2 : Aim: To examine the effect on macroscopic growth of U373 tumors in nude mice of anti-TF antibodies given bolus injection i. v.; 10 mg/kg. after pretherapeutic tu- mor growth has been established.

Mice: 60 homozygotous nu/nu 6 weeks old NMRI males.

Groups: Mice are randomly allocated in four groups of 15 and treated with solutions A, B, or C. Treatment starts when 6 consecutive (daily) measurements show Gompertzian growth. This corresponds to 100-200 mm3 Termination: Treatment lasts until the tumors have grown beyond the maximal size of ap- proximately 1.0 cm37 i. e. no tumor diameter larger than 15 mm or until Gompertzian regrowth has been established by 6 consecutive measurements.

At time of termination tumors from each group are excised for histological and immunochemical evaluation. At a weight loss of > 20% or other objective signs of severe toxicity the animal is terminated Parameters: Tumor size in two orthogonal diameters is recorded daily during the growth phase. The body weight is recorded initially and 2 times per week.

Study 111-3 : Aim: To examine the effect of anti-TF antibodies on growth of intracranial U373 tu- mors in nude mice.

Mice: 60 homozygotous nu/nu 6 weeks old NMRI males.

Tumor: U373 implanted orthotopically in the right hemisphere following standard pro- cedures.

Groups: Mice are randomly allocated in four groups of 15 and treated with solutions A, B, or C.

Termination: Mice with signs of chronic neurological impairment are euthanized.

Data: Survival (i. e. time to neurological impairment) is quantified by Kaplan-Meyer statistics.

Example 4 (Assay 14).

In mouse wherein the TF gene is knocked out and human TF gene is inserted (mTF- KO/hTF-KI mice) a 0.5 mi matrigel plug will be located subcutaneously under the abdominal skin. In the matrigel b-FGF (5 ng) will be incorporated and one week later the formation of new patent vessels in the gel will be quantitated by measuring the content of haemoglobin (angiogenesis). The inhibitory capacity (% inhibition of the haemoglobin content) of the hu- man anti-TF antibodies can be evaluated after single or repeated parenteral administrations of the proteins.

Example 5 (Assay 15).

Gene expression analysis assay for discriminating antibodies, that prevents FVlla binding to TF and antibodies, that prevents FX binding to TF.

In cDNA microarray analyses a specific up-regulation of three genes in BHK-TF cells treated with FVlla has been observed. These include: Fra-1, a gene coding for Fos re- lated antigen 1, d2, a gene encoding a member of the helix-loop-helix class of proteins, and Cyr61 encoding an extracellular matrix signalling protein. The following assay is designed to screen for anti-TF antibodies which prevents FVlla induced up-regulation of Fra-1, Id2 or Cyr61.

Cell culture.

Reagents are purchased from GIBCO-BRL Life Technologies unless otherwise noted.

BHK-TF cells, created as described by Poulsen L. K. et al., J Biol. Chem. 273, 6228- 6232, 1998, are grown in Dulbecco's modified Eagle's medium containing 10% FCS, 100 IU/ml penicillin, and 100 pg/ml streptomycin to obtain 95-100% confluence, washed and grown for additional 16-18 hs in medium without FCS. The cells are again washed and ex- posed to FCS-free medium containing 100 nM FVlla.

For cloning of fragments for Northern blot analyses the cells are treated as follows. BHK-TF cells are grown in Dulbecco's modified Eagle's medium containing 10% FCS, 100 IU/ml penicillin, and 100 Rg/mi streptomycin to obtain 95-100% confluence, washed and grown for additional 16-18 hs in medium without FCS. The cells are again washed and exposed to FCS-free medium containing 100 nM FVlla for 1 h. CRL2091 cells (ATCC) are grown in Is- cove's modified Dulbecco's medium containing 10% FBS, 100 U/ml penicillin, and 100 ug/ml streptomycin to 95-100% confluence. Subsequently, the cells are serum-starved for 16-18 hs and treated with FBS-free medium containing 100 nM FVlla for6 hs. Murine 3T3-L1 cells (ATCC) are maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 pg/ml streptomycin. Cells are grown to conflu- ence and induced to with media containing 1 uM dexamethasone (Sigma), 10 pg/mi human insulin (Novo Nordisk A/S), and 1 uM BRL49653 (Novo Nordisk A/S) for 1 h.

Cloning of fragments for Northern blot analyses.

Fra-1 is cloned by reverse transcription PCR from RNA isolated from 3T3-L1 cells treated for 1 h with dexamethasone, insulin, and BRL49653 using the superscript 11 kit (Life Technolo- gies) according to the manufacturer's instructions. Id2 and Cyr61 are cloned by reverse tran- scription PCR from RNA isolated from BHK-TF cells treated for 1 h with FVlla and from CRL2091 cells treated for 6 hours with FVlla, respectively. The upstream and downstream primers are: 5'-GCGGCCGCCATGTACCGAGACTACGGGGAACCG-3'and 5'- GCGGCCGCTCACAAAGCCAGGAGTGTAGG-3'for Fra-1, 5'- CAGCATGAAAGCCTTCAGTC-3'and 5'-CTCTGGTGATGCAGGCTGAC-3'for Id2, 5'- CGTCACCCTTCTCCACTTGA-3'and 5'-CTTGGTCTTGCTGCATTTCT-3'for Cyr61. Pa- rameters for PCR are one cycle of denaturing at 94 °C for 10 s, annealing at 65 °C for 15 s, and extension at 72 °C for 1.5 min, one cycle of denaturing at 94 °C for 10 s, annealing at 64 °C for 15 s, and extension at 72 °C for 1.5 min, one cycle of denaturing at 94 °C for 10 s, an- nealing at 63 °C for 15 s, and extension at 72 °C for 1.5 min, one cycle of denaturing at 94 °C for 10 s, annealing at 62 °C for 15 s, and extension at 72 °C for 1.5 min, one cycle of dena-

turing at 94 °C for 10 s, annealing at 61 °C for 15 s, and extension at 72 °C for 1.5 min, one cycle of denaturing at 94 °C for 10 s, annealing at 60 °C for 15 s, and extension at 72 °C for 1.5 min, 40 cycles of denaturing at 94 °C for 10 s, annealing at 55 °C for 15 s, and extension at 72 °C for 1.5 min. All fragments are cloned into into TOPO 2.1 (Invitrogen) and sequenced using a Megabase sequencer.

Northern blot analysis.

Total RNA are isolated from BHK-TF cells incubated with FVlla, FX, ASIS, 1 F44A1 or TF8- 5G9 using TriZol following the instructions of the vendor. 20 pg of RNA are size-fractionated in a denaturing gel containing 1 % agarose, 20 mM MOPS, 5 mM NaOAc, 6% formaldehyde, and 1 mM EDTA, transferred to a Hybond N+ membrane (Amersham) by capillary blotting and immobilized by UV crosslinking. cDNA encoding Fra-1, Id2 or Cyr61 are labelled with the Prime It kit (Stratagene) using [a 32p] dATP (Amersham) and hybridized using Express Hyb (Clontech) following the manufacturer's instructions and results are visualized by autoradio- graphy.

Example 6 (Assay 16).

MAPK assay via the Elk1 transcription factor/Luciferase reporter (PathDetect) HeLa cells are seeded to 40 % confluence in a T-80 flask one day prior to transfec- tion. Cells are transfected with 150 ng pFA-Elk1 (Stratagene), 3 jug pFR-Luc (Stratagene), 3 gg human TF/pcDNA3 and 3 ttg mouse Protease Activated Receptor 2/pcDNA3, 1+ using 36 gl FuGene (Roche) as described in the manual. The following day the cells are detached by VerseneTM (Invitrogen) and seeded in black 96 well view plates (Packard) at a cell density of 20.000 cells per well. After the cells had reattached to the plate, the medium is replaced with 160 gel per well serum-free Dulbeccos Modified Eagle Medium (Invitrogen) and incubated for 16 hours.

Cells are preincubated for 1 hour with either 20 lli serum-free medium (control), 20 gi 2, 5 gM FFR-rFVlla (control), 20 pl 2, 5 p, M anti TF antibody B or 20 pl 2, 5 p, M anti TF anti- body A. 20 µl 0,5 µM FVIIa is added to half of the wells and medium to the other half. Follow- ing 4 hours of incubation the cells are subjected to the Luciferase gene assay. LucLite (Packard) reagent is added to the cells as described by the manufacturer. Luciferase expres- sion levels are read on a TopCount Microplate Scintillation (Packard).

Example 7: Preparation of humanized antibodies immunospecific for human TF.

Human monoclonal antibodies are made as described in example 1. These antibod- ies have relatively low affinity to tissue factor compared to murine monoclonal antibodies made against the same epitope. The hyper variable regions of the murine monoclonal anti- bodies have been sequenced and transferred to a human monoclonal antibody that reacts with the same epitope on tissue factor as well as to a"nafve"human monoclonal antibody framework. Affinities have been compared between the two antibody frameworks.

Antibody generation.

Transgenic mice expressing human antibodies and normal mice (e. g. Balb/C or RBF) are three times immunized with the antigen, here human TF. 10 days after the last im- munization the antibody titer were determined and the highest responder of the transgenic mice and the normal mice were selected for fusions. Spleen cells of these mice were isolated and fused to the mouse myeloma cell line AG8. X63 in the case of the transgenic mice and FOX in the case of the normal mice. After 14 days of selection clones were screened for binding towards human TF. Afterwards the affinity of positive clones is measured by BlAcore. mRNA extraction and cDNA synthesis. mRNA from clones of the highest affinity mouse and human antibodies is isolated and used as template for cDNA synthesis by RT-PCR amplification of the hypervariable re- gions according to standard procedures.

Amplification of murine TF antibody variable regions.

TF antibody variable heavy chains are amplified from cDNA using a degenerated primer mix of: 5'-ACTAGTTTTGGCTGAGGAGACGGTGACCGTGG-3', 5'-AC- TAGTTTTGGCTGAGGAGACTGTGAGAGTGG-3', 5'-ACTAGTTTTGGCTGCAGA- GACAGTGACCAGAG-3', 5'-ACTAGTTTTGGCTGAGGAGACGGTGACTGAGG-3'and a universal primer mix present in SMART RACE kit supplied by Clontech (kat. # K1811-1).

TF antibody variable kappa-light chains are amplified from cDNA using 5'-TCAT- CAACACTCATTCCTGTTGAAGCTCTTGA-3'and a universal primer mix present in SMART RACE kit supplied by Clontech (kat. # K1811-1).

Amplification of human TF antibody variable regions.

HuTF antibody variable heavy chains are amplified from cDNA of using primer 5'- GTGCCAGGGGGAAGACCGATGGG-3'and a universal primer mix present in SMART RACE kit supplied by Clontech (kat. # K1811-1).

HuTF antibody variable kappa-light chains are amplified from cDNA using primer 5'- GCAGGCACACAACAGAGGCAGTTCCAGATTTC-3'and a universal primer mix present in SMART RACE kit supplied by Clontech (kat. # K1811-1).

Cloning of TF antibodies into E. coli The PCR product of the TF antibody variable chains are TOPO-cloned according to manufactures protocol (Invitrogen kat. #45-0030) and transformed in competent E. coli cells.

DNA is extracted from transformants and the variable regions are characterized by sequenc- ing.

Genetic modification of HuTF antibody hyper variable regions.

The CDR's are characterized by sequencing from which conserved consensus mo- tifs can be derived. Based on the sequence data obtained from HuTF and murine TF anti- body variable regions new CDRs are engineered and introduced into the HuTF antibody framework. The new CDRs are created in the variable regions by extension PCR and the engineered variable regions are introduced into IgG expression vectors plESRg1, pIESRg1f, and plESRg4 (provided by Medarex) in the following manner: Variable light chains intro- duced into HindllllBsiWi site and variable heavy chains introduced into NotllNhel site of the vectors.

Expression of recombinant TF antibodies.

The resultant plasmids are analyzed by sequencing and transfected into mammalian cells according to standard procedures. Antibodies are purified from the supernatant by pro- tein A and affinities between those recombinant generated antibodies are compared.